[001] This technology relates to belt splicers of the kind disclosed in patent publication US-7,325, 580 (Feb 2008, Woolner+).

[002] In a belt-splicer , the two ends of the belt that are to be spliced are brought together in an overlapping relationship, and appropriate adhesive is applied. Generally, the adhesive has to be vulcanized under conditions of heat and pressure. The belt-splicer includes an operable pressure pad for exerting essu e on the joint, and includes a heater for heating the joint.

[003] The pressure pad structure traditionally includes an inflatable bag. The bag is inflated with fluid. The fluid is air, for pressures up to about six atmospheres , and a liquid at higher

pressures. Often, only one bag is provided, but two bags — one above the belt and one below — are sometimes used. Generally, two heaters are provided, one above and one below the belt (i.e tha joint between the two ends of the belt) .

[004] A conventional belt-splicer includes an upper beam, or an upper set of beams, and includes a lower beam, or lower set of beams. The lengths of the boams span across the width of the bfsilt, above and below. The beams are long enough such that the end3 of the beams protrude out beyond the width of the belt.

[OO'i] The beams are arranged in pairs, the pair comprising on© upper beam and one lower beam. Left and right tension-links are provided in respect of each pair of beams, and the links extend vertically between the respective ends of the beams. The pressure exerted by the inflatable bag, which urges the beams apart, is reacted by tension in the links.

[006] The traditional tension-link has been in the form of a metal rod, having threads of opposite hands on the ends, and respective oppositely-handed threaded nuts. The nuts engage slots in the ends of the beams. The tension-link can be adjusted for length by rotating the rod, as in a turnbuckle.

[007] The tension-links must not be allowed to fail due to the tensile force. The force in each link is typically ten or fifteen tonnes, and the link must be designed to support the force, with an adequate safety margin. [008] One of the problems facing the designer, when specifying the safety margin, is that the linke do suffer abuse. The operation of belt-splicing is carried out in the field, often out in the open, and often at locations that involve climbing ladders, etc. The result is that the beams and the links are subjected to many different and unpredictable abuses, due — not BO much to over-stressing, in

operation, but rather — to violent knocks and similar other abuse, during transport of the equipment to the splicing site, and during handling at the eite.

[009] As an illustration of the unpredictable nature of the abuse suffered by the links, it iet not unknown for the traditional rod-and- nuts link to be used as a handy sledge-hammer. Also, in many cases, the abuse can be habitual and on-going, and it often happens that the fact of the abuse is not recorded. Also, often, although the abuse often does leave marks, there is no indication whether the abuse has been bad enough to call into question the continuing integrity of the link. The abused link is simply used, and then put back into storage, ready for re-uee.

[0010] One consequence of this habitual and on-going abuse, over a period of time, has been that tho links can start to develop cracks. The high tensile stress during operation exacerbates the cracks, and it has not been unknown for the links to fail, it is also not unknown for the links to fail while being subjected to tensile forces that are signi cantly below the designed-for force, it may be noted that the said US-7,325,580 was concerned, not with preventing the links from failing, but with minimizing the consequences that arise from the failure.

[0011] In most engineering applications, a problem of premature failure of a tensile link can be addressed simply by increasing the teneile capacity of the link. In this case, however, the problem cannot be addressed that way. (Of course, increasing the load capacity does no harm — except that it wastefully increases the resources needed to produce the link. )

[0012] Rather, what is required 3 to provide the link with an ability to resist violently-abusive damage, and — even more

importantly — to provide the link with the ability to show when it has been damaged. That is to say, it is important, if the link looks undamaged, that the link should actually be undamaged, i.e that the link has not suffered damage of such nature and magnitude as to affect its integrity. The corollary — that, if the link does show visible signs of damage, that the link actually is damaged — ^■ is leas

important _/ in that it is a simple matter to instigate a program under which, if the link shows visible damage of any kind, it must be discarded — much the same as is done with wire-rope slings, for the same reason.

[0013] it is important to note also that the person who makes the assessment of the fact of, or extent of, damage to the link, on a day- by-day basis, is the splicer-operator , or the storekeeper, rather than a professional safety officer. The mark left by damaging abuse should be "unmlstakeable" in this context.

(0014) Another problem associated with the traditional metal rod- and-nute link used in belt-splicers, is the weight of the links. Such links can weigh 201bs or more, each. The fact La that belt-splicers often have to be deployed in such locations and circumstances that all the components of the belt-splicer have to be carried, by hand, up a ladder. While this ie accepted as a fact of life, designers of belt- splicers are under constant pressure to make the components lighter, and easier to handle.

[0015] it is also the case that, when an operator is carrying things, and is on a ladder, they should at all times have one hand free, to hold onto the ladder or other support. Again, designers have to bear this requirement in mind.

[0016] Some of the components of a splicer are heavy but easy to hold and to carry; some components are light but awkward to carry. The beams, for example, typically are formed as aluminum extrusions, and are of such shape that one beam can easily be picked up and carried, even though large and bulky. The rod-and»nuts links, however, are heavy and awkward to carry. Basically, an operator can only carry one link at a time, up a ladder. Thus, often, many journeys are needed to transport all the components of the belt-splicer to the job site.

[0017] A first aspect of the preeent technology concerns a las ic covering for a tension-link, . the tension link that is used for linking the ends of the transverse beams together, in a belt-splicer, ie encased or encapsulated in a protective covering of a soft plastic material, such as polyurethane. The plastic covering is arranged in such manner that, if the plastic has the visible appearance of being undamaged, the link itself must inevitably be undamaged (in the sense of still retaining its tension-supporting integrity and its designed safety margin) .

[00,18] Thus, if the plastic covering has a visible cut or slit, for example, that fact is immediately apparent to even a casual operator ox- storekeeper, and it is a non-debatable decision on that person's part, to discard every such link. By contrast, if it were the case that the operator had to make a judgment verdict whether this or that mark or contusion of the plastic is or is not damaging, that would not be satisfactory. But it can confidently b© left to operators and

storekeepers to make the determination, for example, whether, or not, there is an actual incision in the plastic.

[0019] in the case of a plastic covering, if the plastic (even though marked and bruised) is intact and unbroken, that condition indicates that the load-carrying structure of the link has not been compromised — with sufficient confidence that the link can be passed for re-uee. Only if the plastic is cut or broken is there a chance that the load-carrying structure might be compromised.

[0020] it may be noted that this favourable no-f lse-positivea situation does not arise in the case of other kinds of load-carrying member, if the manner in which the member la abused is sim le

overloading. In the special case where the load-carrying member is a tension-link in a belt-splicer, the links are very rarely subjected to simple overload. Traditionally, belt-splicer links have been the subject of failures at loads below their theoretical capacity — but tho cause is damage arising from abusive handling, not damage arising from excessive loads.

[0021] Thus, plastic encapsulation of the link, as a way of ensuring the load-carrying integrity of the link, is especially advantageous in the case of a tension-link for a belt-splicer.

[0022] A incision in a plastic covering, ia hugely more visibly apparent than even a largo (and highly dangerous) crack in the traditional steel rod-and-nuts link. Such a crack, if present, inevitably lies hidden among the screw-threads.

[0023] It may be noted that a plastic-encapsulated link is very much less likely to be used as a sledge hammer than the traditional metal rod-and-nuts link — which does at least have a hard heavy head and a rudimentary shaft.

[0024) When the link is encapsulated in plastic, it is a simple matter to configure the moulded plastic so as to form a handle, thus simplifying the task of carrying the links up ladders, etc. AIBO, different sizes, or different load-capacitieB , etc, of link can be Very clearly di ferentiated, simply by using different colours of plastic.

[0025] Preferably, the plastic covering should be so configured that the plastic is not exposed to the heavy tensile forces that are to be supported by the link. Thus, the link (comprising the tension- supporting structure and the plastic covering) should be so configured as to leave metal surfaces exposed, through which the forces from the beams are actually fed into the link. In other wordG, the plastic material should not, itself, be in the load-line.

[0026] preferably, the plastic-encapsulated link should be adjustable as? to its length, for the same reasons a3 the traditional rod-and-nuts link is adjustable. Conveyor belts that have to be spliced are of different thicknesses; and the aggregate thicknesses of the associated equipment (inflation bags, heaters, platens, etc) is not always the same.

[0027] Preferably, the tension-supporting structure of the link is in th© form of an endless loop of rope. The fibres forming the rope can be metal. However, preferably, the fibres are of a strong but light non-metal such as KEVLAR (trademark), poly paraphenylene

terephthalaroide. As described herein, th© rope is in the form of an endless loop, preferably running-track shaped. Preferably, the fibres are not e.g wound around in a circle and then bent to the running-track shape, but the loop is actually manufactured in the running-track shape.

[0028] One of the benefits of using the synthetic rope-loop lies in its failure mode. If the rope-link were to be overstressed (in tension), the rope tends to yield and stretch, whereby the rope-link does not suddenly let go. By contrast, the conventional rod-and-nuts links, when they failed, fractured suddenly. This could be very dangerous, because a good deal of potential energy is stored in the pressurized air bags, and this energy could be released suddenly upon failure of the bolts.

[0029] The plastic encapsulation of the rope-loop protects the rope very effectively against knocks, contusions, abrasions, cuts, and the like. It should be understood that nearly all the abuse to which the links are subject comes during storage-handling and transport-handling, when the link is under no load, rather than during actual load- supporting operation.

[0030] It is recognized that the moulded plastic encapsulates all the vulnerable parts of the tension-link unit, but yet the plastic does not lie in the path or load-line of any of the heavy forces. The plastic can be damaged by knocks, but it ia thick enough to shrug off almost all abuse, apart from cuts to the plastic, which could go right through and damage the fibres of the rope inside.

[0031] it ie recognized that if the plastic is damaged enough that the rope is or might bo compromised, that fact is obvious to an observer.

[0032] Preferably, the length of the tension-link, being the distance apart of the upper and lower load-receiving metal surfaces, is not lass than 20cm and not more than 70cm. Lass than 20cm, the link would be too small for handling to be a problem whatever its structure; above 70cm, handling will be a great problem whatever its structure.

[0033] A second aspect of the technology concerns a tension-link unit which includes couplings that include a capability to pivot.

[0034] The rod-and-nuts links of US-7,325, 580 could be overstressed as a result of the rotation of the ends of the beam as the beam underwent bending deflection, The link as described provides a coupling that allows pivoting at the joint between the link, and the beam. Thus, the described link itself is not subjected to a bending moment, as was the rod-and-nuts link.

[0035] The tension-link unit, as described herein, ia stressed only in tension, during operation. The described link is also significantly stronger than the conventional links of comparable size. The described links are typically less than a quarter of the weight of the

conventional links.

[0036] in the designs described herein, the tension-link is not pivoted directly into the beam. Rather, the tension-link pivots directly relative to the cheek-block of the coupling, and the coupling slides into the beam. The tension-link has to be assembled and disassembled from the beams, every job, and its easier to make that possible by using a slide-in block, than by a design that requires operators to make/break the pivot joint. Besides, preferably the beam is an extrusion, which lends itself to slide-in shapes. Also, the cheek-block, with its flat sides, spreads out the heavy force between the beam and the ten ion-link over a large area of the (soft) aluminum of the beam, and it would be difficult to provide a comparable load- bearing area at a pivoting joint.

LIST OP DRAWINGS, AND DESCRIPTION OF PREFERRED EMBODIMENTS

[0037] The technology will now be further described with reference to the accompanying drawings, in which:

[0038] Fig.l is a pictorial view of a belt-splicer. This design uses three pairs of beams, which are linked together at their ends with plastic-encapsulated tension links.

[0039] Fig.2 is a similar view of a belt-splicer that has just a single-pair of beams. The belt to be spliced, and the heaters, pressure pads, etc, have been omitted from this drawing.

[0040] Fig.3 is an end elevation of a belt-eplicer, showing the manner in which the tension-link units interact with pairs of beams. In Pig.3, spacers have been provided between the beams, which perform the functions described in the said US-7 , 325 , 580.

[0041] Fig.4a is a front elevation that shows a belt-splioer

assembled in place onto the belt.

[0042] Fig. b is the same view as Fig.4a, except that now the

pressure pad has been energized, exerting a large compression pressure on the joint in the belt. The upper and lower beams have deflected in bending mode, whereby the ends of the beams have undergone rotation- The couplings of the tension-link units have pivoted correspondingly .

[0043] Fig.5 is a pictorial view of a tension-link unit, which

includes a tension-link and upper and lower pairs of attached pivoting couplings, by which the tension-link interacts with tho upper and lower beams,

[0044] Figs.6a, 6b are pictorial sectioned cut-away views, showing the structure of the tension-link,

[0045] is close-up, sectioned side-view, showing some details of the tension-link unit, and of its

interaction with one of the beams.

[0046] Fig.8 is a plan view of a rope that has been formed into a running-track-shaped endless-loop, the fibres of the rope having been wound round spaced formers.

[0047] Pigs. a, 9b show a tension-member, comprising the endless- loop of rope shown in Pig.8 and two bosses that have been assembled into the rope.

[0048] Figs.10a, 10b are a plan view and a sectioned side-view of a mould-half, in which the plastic encapsulation is to be carried out,

[0049] Figs.11a, lib are a plan view and a sectioned side-view of the mould-half, into which the tens on-member shown in Figs. a, 9b has been placed.

[0050] Fig ,12a is the same as Fig.11a, except that, now, mandrels placed in the bosses are being forced apart. This produces tension in the fibres of the rope.

[0051] Pig.12b is a sectioned side-view corresponding to Fig.12a, except that the other half of the mould is now included.

[0Q->2] Fig.13 is a front elevation of a portion of a belt-splicer that is to be dismantled. The left and right tension- link units are in place, operationally engaged in the ends of the upper and lower beams .

[0053] Pig.13a is the same view, but shows only the ends of the beams .

[0054] Fig.14 shows the situation when the upper coupling of the tension-link unit has been slid out of the slots in the upper beam. The unit has been rotated about the pivot axis of the lower coupling, and lies at an angle to the vertical, and the lower coupling is still engaged with the lower beam.

[0055] Fig.15 ehows the situation when the unit has been pivoted some more, about the lower pivot axis,

[0056] Fig.16 shows the angle of the tension-link now aligned with the length of the lower beam. Now, the lower coupling has been slid further into the lower beam. The tension-link and the upper coupling also have followed the lower coupling into the hollow interior of the lower beam.

[0057] Pig.17 is a part-sectioned side-view of a pivot-pin and a pair of cheek-blocks of a tension-link unit.

[0058] Fig.18 shows the same structure in front elevation, and shows the eccentric location of the position of the pivot axis in the cheek-block.

[0059] Fig.19 ie a diagram that shows the various configurations into which e couplings of the tension-link unit can be rotated, thereby enabling the distance between the cheek-blocks of the tension-link unit to be adjusted.

[0060] The scope of the patent protection sought herein is defined by the accompanying claims. The apparatuses and procedures shown in the accompanying drawings and described herein are examples.

[ 0061] .Some of the physical features of the apparatuses depicted herein have been depicted in just one apparatus. That is to say, not all options have been depicted of all the variants. Skilled designers should understand the intent that depicted features can be included or substituted optionally in others of the depicted apparatuses, where that is possible.

[0062] Some of the components and features in the drawings have been given numerals with letter suffixes, which indicate upper/lower etc versions of the components. The numeral without the suffix has been used herein to indicate the components generically.

[0063] Terms of orientation (e.g "upper/lower" , "left/right", and the like) when used herein are intended to be construed as follows. The terms being applied to a device, that device is distinguished by the terms of orientation only if there is not one single orientation into which the device, or an image (including a mirror image) of the device, could be placed, in which the terms could be applied

consistently.

[0064] Terms used herein that define respective theoretical constructs, are intended to be construed according to the purposive construction.

[0065] F.ig.1 shows two ends 21a, 21b of a belt, which are to b© spliced together. The belt-splicer 23 includes heater/cooler platens 25U,25L, and also pressurizable airbags or pressure pads 27. To splice the two ends of the belt together, suitable adhesive is applied to the joint, then the pressure-pads are energized to press the ends together, and heat is applied. The heat and pressure are maintained for a period of time, in order to vulcanize /cure the adhesive. [0066] When carrying out a splicing operation, typically all the components of the belt-splicer must be taken to the site where the splice ΐθ to be made. Often, this has to be done by hand. Often, climbing of ladders, while carrying the components, is involved.

Therefore, the components should be light and easy to carry. he designers should have it in mind that the components will b© dropped and knocked, and otherwise abused, and therefor© general robustness is desirable. The components should be light enough to be manhandled into position, and should be easy to assemble and disassemble. Of course, the use of tools is not ruled out, but the fewer tools, nuts and bolts, etc, and the like, that have to be carried to the splice site, the better .

[0067] The present technology is concerned with the interactions between the beams, being the upper and lower beams 29U,29L that span across the belt 21, with the tension-link units 30 that link the ends of the beams 29 together, in the drawings, in preparation for splicing, the eams 29 are manoeuvred into place above and below the belt 21, and the tension-link units 30 are engaged into the ends of the beams. Stop-pins 32 (see e.g Fig.13a) are used to keep the tension- link units 30 in their proper places, in the ends of the eams, once assembled .

[0068] he beams 29 are formed as aluminum extrusions. The extruded profile of the beam defines an enclosed hollow space 34. The space 34 includes recesses or slots 36 in the side-walls 38 of the beam. The slots 36 are formed with respective force-transmitting surfaces 40 — being the surfaces at the bottoms of the slots 36 in the case of the upper beams 29U, and the surfaces at the tops of the slots 36 in the case of the lower beams 29L. The force-transmitting surfaces 40 are the surfaces that make forceful contact with the engagement- faces 41 of the tension-link units 30, when the pressure pad 27 is energized, and the tension-link units are performing their role of holding the upper and lower beams 29U,29L together.

[0069] Pig.4a shows the assembled belt-splicer 23, with the ends

21a, of the belt having been prepared, and adhesive applied, and the splicer now ready for the pressure pad 27 to be inflated. Fig.4b shows the pressure pad having been inflated. Now the upper and lower beams 29U,29L have deflected in bending; this has caused the ends of the beams to rotate through an angle. The tension-link units 30 are provided with pivotable cheek-blocks 43, which can pivot and so follow the rotations of the ends of the beam, thereby preventing any perverse bending moments from being transmitted to the tension-link unit3 30.

[0070] In the belt-splicers that used the traditional rod-and-nuts tension-links (see US-7,325,580) , the links, there, were indeed subjected to perverse bending moments that were induced by the rotation of the ends of the beam as the beams bent under load. The elimination of such bending moments in the present tension-links is considered a significant safeguard against prematures failure.

[0071] The structure of the tension-link unit 30 is shown in

Figs.5-7. The tensile strength of the tension-link unit 30, from the standpoint of supporting the forces tending to sepa ate the beams

29U,29L, is determined by an endless-loop 45 of rope, which passes round upper and lower bosses 47U,47L. The bosses 47 are shaped like pulleys; it is (stressed, however, that the bosses 47 are not intended to rotate, or to undergo any motion at all, relative to rope 45.

[0072] The sub-assembly shown in Fig.5 is referred to herein as a tension-link unit, designated 30. Th© device shown in Figs.6a, 6b is referred to herein as a tension-link, designated 49» The tension-link unit 30 in Fig.5 comprises the tension-link 49 and two couplings 50. The coupling 50 comprises the boss 47 and the pivot-pin 52. The pivot- pin 52 engages a through-hole 54 in the boss 47.

{0073] The pivot-pin 52 also carries the cheek-blocks 43. In the illustrated structure, th© cheek-blocks 43 are used in pairs; both cheek-blocks of the pair are locked to the pivot-pin, such that both cheek-blocks are forced to rotate in unison with th© pivot-pin.

[0074] (Preferably, the cheek-blocks 43 should not be locked tightly to the pivot-pin 52. In Pig.17, for example, the grub-screw 56 is tightened into the cheek-block, but is a loose fit in its recess in the pivot-pin 52. The heavy tensile forces that are applied to the tension-link unit 30 pass through the interface between the pivot-pin 52 and the through-hole 54, and the mentioned looseness permits the pivot-pin to "settle into" the through-hole 54 in the block 43, under this heavy force, whatever the rotational orientation of the pivot-pin + cheek-blocks relative to the boss 47.)

[0075] The tension-link 49, as shown in Fig.5, comprises the endless-loop 43 of rope, the two bosses 47U,47L, and a plastic covering 56. The loop 45 of ropo passes around th© upper and lower bosses

47U,47L, whereby the loop is running-track-shaped, comprising two semi- circular uns connected by two straight runs.

[0076] It is recognized that the plastic covering does not tend to break, free from, and become detached from, the fibres of the rope, under heavy loading of the tension-link. Although the rope does stretch under heavy tension, the plastic io pliable enough to follow the motion, and "breaking free" is not a problem. In this regard, preferably the rope should be under an applied tension while the liquid plastic is being poured over and around the rope, during moulding, and during curing of the plastic,

[0077] It is recognized that that applied tension, during moulding, need not be the full maximum tension for which the tension-link is rated; but the applied tension should be enough at least to take up the slack in the rope and in the fibres, so that at the moment the fibres are encapsulated, the fibres are, more or less, in the positions they will be in when the heavy force comes on. It is recognized that any slight further movements of the fibres relative to each other, as between the slack-taken-up condition and the fully-loaded condition, are small enough not to affect the integrity of the moulded bond between the fibres and the plastic covering,

[0078] In the straight-runs of the endless-loop 45, the cross- soctional profile of the rope will be the rope's own inherent shape. However, in the semi-circular-runs, over the bosses 47, the profile of the rope will follow the profile of the groove provided in the outer sux' ace of the boss 47, In Pig.7, the groove has a round circular profile. Thus, the fibres of the rope, as they go over the boss, press deeply into this groove, and adopt the shape of the groove.

[0079] The endless loop of rope is prepared by winding the fibres over the bosses (or over formers that duplicate the profile of the bosses), whereby the cross-sectional profile or shape of the rope, as it passes over the boss, is as shown in Fig.7. It should be understood that, in Pig.7, the loop of rope has been done in such a way that the fibres are not bent over the boss after the rope was formed— which would have caused the outer fibres of the profile of the rops to be tight, and the inner fibres to be slack. Rather, when the rope was being formed to the shape shown in Pig, 7, the rope was wound under tension. Thus, all the fibres, in all locations in the profile, are « at least notionally — under the same tension. (By contrast, usually when a loop of rope is bent over a pulley, under tension, of course the outer fibres of the rope are then under much more tension than the inner fibres . )

[0080] The groove in the boss 47 is circula , but the mentioned condition — that all the fibres are under the same tension — applies whatever the profile of the groove in the boss, provided the rope has been formed by winding the fibres around the two bosses, under tension. That is to say; this manner of forming the rope leaves each fibre the length it needs to be, to be at the same tension as the other fibres.

[0081] The groove should be a channel that has side-walls high enough to contain all the fibres, and the groove surface should be smooth and well radiused, but, apart from that, the grooves might be e.g flat-bottomed, or any shape.

[0082] It is not required that the groove in the upper boss 47U be the same sha e as the groove in the lower boss AIL, although same-ehape is preferred. It is not essential that the semi-circular rune be strictly of a constant radius; however, constant-radius-circular is preferred .

[0083] it may be noted that the bosses 47 are not fixed or locked to the loop 45 of rope, nor to each other, nor into the tension link 49 as a sub-assembly, other than by being moulded into the plastic covering 56. Th s, if a heavy turning moment were to be applied to the boss 47, the boss would turn and the plastic covering would be damaged. It is recognized that the presence of the couplings!, with their pivoting capability, isolates the bosses from all but small

insignificant turning moments.

[008 ] Thus it is recognized that, in the illustrated design, the bosses and the loop of rope do not need to be held in their mutual locations by anything more than the plastic covering — because the perverse forces and turning moments that might tend to move and dislodge the components happen during handling (including the

inevitable rough handling), and do not happen when the tension-link is under heavy operational loading, it is recognized that the plastic covering, as described, is able to provide an excellent service life performing this (limited) role, and that the bosses need not be otherwise fixed to the loop of rope.

[0085] The manufacture of the tension-link is shown in Fig8.8-12b.

In Pig.8, the endless-loop 45 of rope has been formed by winding the rope fibres around formers. In Fig, 9, the formers have been removed and replaced by the bosses 47U,47L— or the rope could have been formed around the bosses themselves. In Figs.9a, 9b, mandrels 58 have been placed in the through-holes 54 in the bosses .

[0086] Figs.10a, 10b illustrate a mould-half 60. Figs.11a, lib show the components of Figs.9a, 9b now placed in the mould-half 60. The mandrels 58 protrude through right and left holes in the mould-half, it will be noted that the left-side hole is elongated, in

Fig3.12a , 12b, a force has been applied to the mandrels, urging them apart, as shown by the arrows. The elongated hole means that the endless loop 45 of rope can stretch, as a result of the force applied to the mandrels 58 In Fig.12b, the mould has been completed, and the pouring of the liquid plastic takes place while the rope is being stretched. Preferably, the tension is maintained also during curing of the plastic.

[0087] Preferably, the plastic is cold-pouring polyurethane. In- mould curing of the plastic is done in an oven set to the appropriate temperature, hen the plastic material is poured, it is in liquid form. It is important, from the standpoint of the quality of the plastic that, if any air was introduced when the liquid was being prepared, such air be removed. The liquid should be de-aerated in a vacuum chamber, prior to pouring.

[0088] The mould should be so arranged that the mould seals against the side-faces of the bosses 47, Thus, there is no plastic covering over the side-faces. Also, the through-holes 54 in the bosses, which serve as load-receiving metal surfaces of the tension-link, through which is transmitted the whole tension that is supported by the tension-link, in operation, should be left free of plastic. The mould should be so arranged that, apart from those surfaces, both bosses, and the loop of rope, should be completely encapsulated by the plastic covering .

[0089] A slot can be moulded into the plastic material, to serve as a convenient handle.

[0090] Theoretically, the tension applied during moulding should be no less than the maximum tension likely to be encountered during operation. However, it has been found that, so long as the tension applied during moulding is enough to take up the slack, applying further tension during moulding has little benefit. It has been found that the first tension applied to the loop of rope, wound round the bosses, takes up the slack at a fairly low rate — that is to say, an incremental increase in forces produces a large elongation of the loop. But once the slack has been taken up, the rate changes, and now an incremental increase in force produces only a tiny elongation. Thus, the tension in the loop should be enough to take up the slack.

[0091] inevitably, when the load goes on, the fibres roust undergo some movement relative to each other. However, when the fibres are wound, under tension, around formers that have the same profile and spacing as the bosses, as described, the amount of such relative movement of ^" the fibres, during operation, is minimized. It is recognized that, if slack-take-up tension is applied to the loop of rope as the liquid plastic is poured into the mould, and during curing, there is no problem of the plastic separating from the fibres during operation.

[0092] The tension-link, as described herein, comprises the loop of rope and the two bosses, encapsulated in the moulded plastic covering. The tension-link, unit comprises the tension-link, plus the upper and lower couplings. Each coupling comprises a pair of cheek-blocks, and a pivot-pin. In the tension-link unit, in respect of each coupling, the pivot pin is assembled into the through-hole in the boss, and the two cheek-blocks are locked to the pivot pin, one each side of the boss. The pair of choek-blockc is rotatablo, as a unit, with respect to the boss ,

[0093] Preferably, the couplings are factory-assembled to the tension-links, whereby the tension-link units are shipped, ready- assembled, with the two pairs of cheek-blocks captivaly assembled to the tension-links. The cheek-blocks are assembled onto the ends of the pivot-pins, and are secured in place with grubscrews. If required, the cheek-blocks can be removed (upon slackening the grubscrews) but tha intent is that they remain attached to their respective rope-links for their service lives.

[0094] Figs.13-16 illustrate another aspect, of the invention, in which the tension-link unit is stored inside the hollow interior of the beam.

[0095] Pig.13 shows a belt-splicer 23 that has been used to create a splice, and now needs to be dismantled, and transported back to the shop. Fig.13a i3 a close-up of one end of the belt-splicer. The tension-link unit 30 is located in cut-outs 65 in the ends of the beams. (The cut-out 65 is only needed on the sides of the eam that faces the other beam, but cut-outs are provided in both sides of the beam so the operator does not have to worry about the beam being the right way up, ) The cut-out 65 defines how far the tension-link unit can be inse ted lengthwise into the length of the beam, when the unit is oriented for operational use.

[0096 J During assembly of the belt-splicer 23, stop-pins 32 were inserted in suitably-located holes in the walls of the beams, which serve to keep the unit in plac© during operation. It may be noted that, in operation of the belt-splicer, the tension-link unit does not engage, or does not forcefully engage, either the cut-out or the stop- pins ,

[0097] Opon disassembly, the* stop-pine 32 are withdrawn (Pig.14).

Now, the tension-link unit can be eased out of the end of the beam, by sliding the cheek-blocks 43 lengthwise along the slots 36 of the beams.

[0098] In the illustrated design, rather than removing the tension- link unit completely from the beams, the unit is separated from one of the eams, and is then inserted, as a unit, lengthwise into the other beam. Fig.14 shows the upper coupling 50U of the unit being withdrawn from the u per beam 29U. The tension-link is rotated about the pivot- axis of the lower coupling 50L, in order to achieve the movements shown in Figs.13-16. It will be understood that the operator manipulates the unit 30 by hand, at this time. t0099] in Fig.15, the tension-link unit has been further rotated, and is now lying nearly flat, and nearly ready to be inserted into the hollow space 34 within the profile of the beam. In Pig.16, this condition has been achieved. The operator orients the cheek-blocks of the upper-coupling 50u, so that the cheek-blocks will enter the slots 36 of the lower beam 29L. Then, the operator slides the whole tension- link unit .lengthwise along the length of the lower beam. It will be noted that the cheek blocks of the lower coupling SOL never leave the slots 36 in the lower beam 29L, so the operator does not need to handle, or set the orientation of, the lower cheek-blocka.

[0100] with the tension-link unit now residing inside the lower beam (Fig.16), the operator can install stop-pins in suitable pre- located holes in the beam, to keep it in place. The tension-link units of that pair of beams can be placed one in each end of one beam, or they can be placed one in each of the beams. [0101] It will be understood that, for die-assembly, the tension- link units are, or can be, placed into their protected storage location inside the hollow beams without even being removed from the beam, Α 3θ, this placement is done before the components even start to be removed from the job-site. Equally, for assembly, the tension-links units are not extracted from their protected chambers until the very lasjt moment, after everything has been transported to the job-site and has been put in position. It is not possible to say that storago-and- handling damage to the tension-link units has been completely

eliminated — but it is almost possible.

[0102] In order for the above-described beneficial manner of storage of the tension-link units to be engineered, of course the units have to b© of a shape and size such that they will fit inside the beam. But, in this case, the units are automatically of such shape and size. The whole of the tension-link unit mU3t fit into the hollow space — but of course it does ao.

[0103] Figs.17-19 embody another aspect of the invention, in which eccentrically-mounted cheek-blocks provide the tension link units with a capability for adjustment. In the previous drawings, the axis of the pivot-pin 52 has been located in the centre of the cheek-block, In Figs.17-19, the axis of the pivot-pin is eccentrically placed in the cheek-block. This enables the tension-link unit to be adjustable as to the distance of separation of the engagement-faces of the upper and lower couplings. The distance apart of the respective engagement-faces of the upper and lower couplings may be regarded as the effective operational length of the tension-link unit. This engagement length is designated J5L.

[0104] Each coupling includes a pair of cheek-blocks, which are arranged to be rotatabla in unison with each other about the pivot axis of the coupling.

[0105] The cheek-block of the coupling is square, and has four faces that define its circumference. These four faces are designated FP, Q , FR, FS. Fig.18 shows the four different radial distances, designated DP, DQ, DR, DS, from these faces to the axis of the pivot- pin 52. DP is the shortest radius, then DQ, then DR, and DS is the longest radius. The length of the side of the square block being L, DP + DS = L, and DQ + DR - L. [0106] The engagement-length EL of the tension-link unit can bo changed /adjusted by rotating the pair of cheek-blocks of the upper coupling and/or by rotating the pair of cheek-blocks of the lower coupling. Rotating one or both couplings, relative to the length of th© tension-link, changes which ones of the faces FP, FQ, FR, FS of the couplings become the engagement-faces of the couplings, and thus changes the length EL of the unit,

[0107] The engagement-face 41U of the upper coupling is whichever of th© faces FP, FQ, FR, FS of the upper-coupling happens to be facing downwards. The engagement-face 41L of the lower coupling is whichever of the faces FP, FO, FR, FS of the lowe -coupling happens to be facing upwards, if/when the cheak-blocks are rotated, the distance from the engagement-face to the pivot axis changes.

[0108] The distance apart of the pivot axis of the upper coupling from the pivot-axis of the lower coupling is designated the distance PAD. The distance PAD does not change, of course, when the cheek- blocks are rotated, but remains constant. The maximum engagement- length ELmax of th© tension-link unit occurs when the face FP of the upper coupling faces downwards, and the face FP of the lower coupling facos upwards. The minimum engagement-length ELmin occurs when the face FS of the upper coupling faces downwards, and the face FS of the lower coupling faces upwards. Dimensionally, ELmax = PAD - 2xFS, while ELmin PAD - 2xFP.

[0109] Other combinations of orientations of the couplings provide intermediate values of the engagement-length EL. The length L of the side of the square block can be notionally divided into twelve unite. A preferred placement of the pivot axis is one in which DP = 4 units, DQ ^w 5 , DR = 7 , and DS - 8 units.

[0110] At this, given a PAD dimension (the unchanging distance apart of the upper and lower pivot-axee of ths tension-link) of e.g 100 unite, ELmax is 100 - 2x4 = 92 unites, while ELmin - 100 - 2x8 - 84 units.

[0111] The seven unit-intervals between these two can all be created by suitably rotating the cheek-blocks, e.g in the manner as illustrated in Fig.19.

[0112] in order to take best advantage of the adjuetabilityprovided by the eccentric blocks, preferably the operators should be given assistance in the form of visible indicia on the blocks, to indicate which orientation the blocks currently stand at, and to simplify the estimation of what rotations to perform to make a large or a small increment of adjustment to the engagement-length EL. preferably, these indicia should go further then marel marking e.g A,B,c,D on the four faces! of the blocks, and should indicate the differences in actual distance. Thus, the blocks should be marked with 4,5,7,8, Reading the marks showing on the cheek-blocks of the upper coupling, and the marks showing on the lower coupling, the operator simply adds them together, to reveal the distance by which the link unit has been shortened, given thoa© orientations,

[0113] Thus, in Fig.19, the mark "1" could be replaced by "4"; the mark "II" by "5", the mark "III" by "7", and the mark "ΙΪΙ1" by "8".

[0114] It will bo noted, in Fig.19, that, at some combinations of orientations of the blocks, the tension-link lies at an angle to the vertical. This is usually of no consequence.

[0115] in order to make the adjustment, of course the coupling has to be withdrawn from its profiled 3lot in the beam. However, this is not a problem. Usually, a splicing job requires the splicer to have several pairs of beams, and the adjustments are done one link-unit at a time. Thus, there would not be a time when the beams were unsupported, as a set.

[0116] it is a trivially simple matter, when the faces of the blocks are suitably marked, for the operators to ensure that all the link-units are set to the same distance. In fact, a good discipline for the operators to follow is to photograph the set of links (i.e two photos, one from each side of the belt), after ad ustment, with all the indicia showing, as conclusive evidence that the failure, if a failure occurred, was not caused by improper adjustment. Such evidence is trivially easy to provide with the tension-link units as described herein, but would have been too difficult with the traditional rod-and- nuts links.

[0117] When the tension-link units are being assembled, in the factory, the cheek-blocks should be assembled to bosses of the tension- link such that, in respect of all four of tho blocks, all the numbers always show upright .

[0118] The scope of the patent protection sought herein is defined by the accompanying claims. The apparatuses and procedures th© accompanying drawings and deecribed herein are examples.

[0119] The numerals used in the drawings can be summarized as»

21a, b (ende of) belt

23 belt-gplicor

25 heater platen

27 inflatable pressure pad

29U,L upper beam, lower beam

30 tension-link unit = tension-link + couplings

32 stop-pin

34 hollow s ace in beam extrusion

36 slot/recess in side-wall of beam

38 3ide-wall of beam

40 force- ransmitting surface of beam, in elot/receas

41U,L engagement-face of cheek-block of coupling of tension-link unit

43 cheek-block of coupling

45 endless-loop of rope

47U,L u er boss, lower boss

49 tension-link

50U/L upper and lower couplings; coupling = cheek-block + pivot-pin

52 pivot-pin

54 through-hole in boas

56 plastic covering of tension-link

58 mandrels, used in the plastic mould

60 mould-half

61 elongated left-side hole in mould-half

3U,L upper and lower load-receiving metal surfaces

65 cut-out in end of beam